Why Banks need a Cloud Strategy for their COTS applications?

Most Banks have started adopting cloud to build its digital enterprise and has become a necessity to stay in the game and to compete with other new age Fintech companies. Within Banks, there are lot of applications that are built by segment leaders, who are better attuned to the industry needs. Over the years large most banks have invested a lot in COTS applications to fulfill their business needs as they provide variety of business functions out of the “box” spanning a wide range of functionality, technology and time-period. While products cover most of functionality needed for specific business function, there are always edge cases that need to be built in-house. Some of COTS applications typically need heavy infrastructure to be provisioned that can be quite resource and maintenance intensive.

With stiff competition setting in the industry from Fintechs and big technology companies and the increasing cloud adoption across the financial services industry, it is high time for banks to relook at their strategy of COTS portfolio they are required to maintain to future proof their growing business needs with reduced overheads. However, COTS applications bring in their own set of challenges with respect to cloud migration.  A typical R path analysis won’t be suitable for defining COTS cloud adoption journey.

Some of the key drivers for banks migrating COTS to cloud

• to increase business agility through automation for provisioning and continuous integration
• to reduce capital investment in infrastructure and licensing cost
• to increase processing capacity utilizing elasticity & scalability of cloud platform
• to address End of Support of COTS products and their underlying technology
• to address Datacenter consolidation and exit plans
• to reduce maintenance overheads due to tools and people required for managing complex environments
• to increase resiliency by utilizing cloud services for logging, monitoring, and scalability

COTS product migration can typically adopt a 4-phase approach as follows:

1. Assess   - analyze product installation, configuration details, infrastructure, security, and data requirements, application affinity (interfaces, shared middleware & data storage)
2. Design   - identify compatible cloud services, components for each of the application, data and integration layers and define target architecture blueprint along with the required migration tools and tasks to perform migration and infrastructure capacity and availability requirements
3. Migrate – execute agile based migration as per design specifications starting with data, integrate with on-prem systems and perform end to end testing including performance and security testing as a pilot with non-prod environment first to ensure no functional impact
4. Optimize – continuous calibration of cloud service configurations to optimize COTS resources as per business needs

 Additionally, the following can help banks in achieving increased business value of migrating COTS workloads to Cloud by:

• automating cloud infrastructure provisioning and tearing down through Infrastructure-as-Code based on demand
• automating operational actions in order to deliver systems that auto-scale & self-heal
• adopting agile development practices: DevOps, Test automation & CI/CD
• adopting cloud-native architectural patterns – stateless, microservices, serverless
• reducing/eliminating licensed software where feasible replaced with open-source software

Some of the key considerations when planning for migrating COTS workload to Cloud are:

• Evaluate SaaS options if organization intends to keep away from future maintenance overheads and application data classification & availability requirements are at expected level from SaaS vendor

• Rehost first to Cloud before attempting replatforming.

• Replatforming of COTS is required when support is not available for original product’s OS/application runtimes/database, non-functional requirements are not fulfilled.  Adopting open-source databases provides savings in licensing cost provided application vendor supports it.

• Check support of COTS applications
  • operating on container platforms
  • dependencies on specific hardware
  • certified for the chosen cloud platforms
  • certified for application technology stack and version
  • data compliance adhering to regional regulators & standards

• Migration to IaaS
  • OS / App-Server / Database / Runtime libraries - upgrade to product vendor supported versions on Cloud
  • Core Application software – obtain new application packages compatible for OS & application server versions from product vendor

Some of the challenges faced when migrating COTS applications to Cloud are:

• products have licensing and dependency on specific hardware are difficult to migrate  
• Latency issues experienced in migrated products having real-time interfaces with on-premises systems 
• products that are running on legacy platforms and yet to be certified by vendor for target cloud platform
• products using shared storage and multicast cannot be migrated to cloud

Migrating COTS applications to cloud is not a straight-forward journey given the wide variety of functional and technical complexities concealed as a black-box. While most large product vendors have cloud-based solutions, upgrading to latest cloud product suite comes with a huge effort when customers are still running older versions. On the contrary, there is a plethora of products still not certified on the cloud and banks have started migrating workloads to cloud. While waiting for certified products on the cloud could be an option, it is equally important for banks customers to look at options of partnering with COTS vendors and Systems Integration partners and start their COTS migration journey to stay in the game ahead of their competition.